Executive Summary
vStore is an approach to enable context-aware storage decisions for applications. It provides a storage abstraction layer with the support for various kinds of storage nodes, ranging from cloud to edge storage.
vStore decouples user-generated data from pre-defined storage locations and hence
- Captures user intentions by combining various sources of contextual information
- Facilitates sharing of data between different applications
- Adapt to changes in the network environment, e.g., in case of congestions.
Contextual Data
A context aggregator provides an interface to different kinds of contextual information. To gather contextual information from the mobile phones, we rely on three providers of such information: First, we make use of AWARE 2 , an open source framework for context instrumentation on Android phones. Second, the Google Places API provides a list of places that surround the user, their type and the likelihood of the users being located at those places. Third, the Android Connectivity API provides information about the network connectivity of the device.
Furthermore, the context aggregator provides an open interface to be extended.
Placement Decisions
vStore uses rules that follow the Event-Condition-Action (ECA) paradigm to make placement decisions. Rules can be defined globally and pushed to the devices, as well as configured locally. A rule specifies certain conditions that have to be fulfilled in order for that rule to be triggered. These can for instance be related to the current contextual information.
Publication
vStore was awarded “Best Paper” at . Mobicase 2018
Gedeon, Julien ; Himmelmann, Nicolás ; Felka, Patrick ; Herrlich, Fabian ; Stein, Michael ; Mühlhäuser, Max (2018): vStore: A Context-Aware Framework for Mobile Micro-Storage at the Edge.
In: Proceedings of the International Conference on Mobile Computing, Applications and Services (MobiCASE), p. 165-182, DOI: . 10.1007/978-3-319-90740-6